Type here Type here Type here [618698]

[Type here] [Type here] [Type here]

Student: [anonimizat] 2013
A Regression Analysis on the Relationship of Final Consumption
Expenditure (EURO) with GDP (million euros)

Sava Maxim Group: 1737 A.S.E. 2017
1
Econometrics

A Regression Analysis on the Relationship of Final Co nsumption
Expenditure with GDP

Abstract
This paper presents the results of the researches regarding GDP analysis in correlation with
the final consumption, an important macroeconomic indicator. The Gross Domestic Product
is highly influenced by the evolution of the final consumption. To analyse the correlation, the
paper proposes the use of the linear regression model, as one of the most appropriate
instruments for such scientific approach. The regression model described in the pap er uses
the GDP as resultant variable and the final consumption as factorial variable.
Keywords: consumption; model; evolution; variable; economy.
JEL Classification: C23, H55, F22

1. INTRODUCTION
Consumption is basic economic activity that individuals encou nter in their everyday lives. It is
actually one of the vital economic activities in an economy. Goods and services indifferent forms
are consumed daily to be able to meet the insatiable wants and needs of people in an economy.
Expenditure on consumption s ignificantly contributes as a portion of the Gross Domestic Product
that is also known as the income of the country. In Macroeconomics, achieving long -term
economic growth is usually the priority of policy makers. Since consumption expenditure is
synonymou s to total spending, consumption is established to be very influential in evaluating the
growth of a country. As Keynes (1936) said, consumption is a function of national income. Through
the years, the government has implemented different sorts of policies that would increase
consumption of the country. These policies are called fiscal and monetary policies. Researchers
and economists have also conducted many studies regarding the importance of consumption in the
economic expansion.
Now that it is establis hed that consumption plays a significant role, this study focuses on the
factors that influence the consumption expenditure in an economy. Firstly, as an aggregate
measure of production, GDP is equal to the sum of the gross value added of all resident
institutional units engaged in production, plus any taxes on products and minus any subsidies on
products. This is a good indicator in evaluat ing the performance of a country mainly because it is
the account of national income (Kuznet, 1946) . Intuitively, there is a negative relatio nship between
the two variables however, saving can be made to be able to consume more goods in the future.
Irving Fisher clarified that consumers are rational and forward looking, making them make decision
that are intertemporal choices which involves taking into consideration different periods of
time(Mankiw, 2010). Third, net taxes on products will be studied. As we know, goods that are
levied with hig h tax can result to a lower demand for that good.

2. REVIEW OF RELATED LITERATURE
2.1. Final Consumption Expenditure
Mankiw (2010) stated that consumption is made up of the goods and services bought by
households. It is made up of three subcategories which are nondurable goods, durable goods, and

Sava Maxim Group: 1737 A.S.E. 2017
2
Econometrics services. Goods such as food and clothing, goods that temporarily exist, are called nondurable
goods. Goods such as cars and computers are called durable goods. These goods last longe r.
Another subcategory is called services. Haircuts, dentist visits and spas are examples of services.
Consumption is part of the GDP equation which will be discussed in the latter part of the paper
under GDP per capita. To explain this even more, the figu re below explains the circular flow
diagram.

Table 1. Circular flow diagram of the capital.

Source: http://1060.edu.pinggu.com/forum/201305/16/1507263og318fkbf20tz8b.jpg

Here, the reader is able to address the circular consumption expenditure of households,
government and firms. The figure shows the flow of received income of the households, how the
income is used to pay for taxes to the government, to avail of goods and services, and to save for
financial markets where households can invest in. Households may also borrow money from the
financial market through the form of loans. The sale of goods and se rvices serve as the revenue of
the firms. The revenue is then used to produce for more goods and services. Taxes serve as the
revenue of the government. These are used to pay for government expenditures. The excess is
then identified as pubic saving.
One m ore important thing to be discussed is the theoretical model of consumption as presented by
Keynes (1936), Friedman (1956), and Modigliani (1948). Keynes started the modern consumption
theory where he analysed the psychological foundation of consumption behaviour in his “General
Theory”. Keynes stated that,

Sava Maxim Group: 1737 A.S.E. 2017
3
Econometrics “The fundamental psychological law, upon which we are entitled to depend with great
confidence both a priori and from our knowledge of human nature and from the detailed
facts of experience, is that men are disposed, as a rule and on the average, to increase
their consumption, as their income increases, but not by as much as the increase in their
income(The General Theory, 1936, p.96). ”
The marginal propensity to consume (MPC) decreases with income. This has the same scenario
with average propensity to consume (APC). However, Kuznet (1946) found that the Keynesian
consumption theory is inconsistent with the data he had gathered about the economy of the United
States. His study showed that constant aggrega te APC characterizes the long run time series
consumption of the U.S economy. In response to the study conducted by Kuznet, Milton Friedman
(1956) came up with permanent income hypothesis (PIH). He explained that households spend a
fixed portion of their p ermanent income on consumption. He described permanent income s the
annuity value of lifetime income and wealth. From this, Friedman came up with a new consumption
function:
𝐶𝑡= 𝑐𝑌∗𝑡
where C= consumption spending, c=MPC, and Y*= permanent income. Modigliani’s theory stated
that MPC is constant and equal to the APC. While Friedman’s theory was being developed,
Modigliani was developing his lifecycle theory at the same time. According to this theory,
individuals choose a lifetime pattern of consumpti on that is optimal in their lifetime utility subject to
their lifetime budget constraint.

GDP
Gross Domestic Product is the measurement of final goods and services. It is essentially the sum
of consumption (C) ,investment (I), government expenditure (G), and net exports (NX). The
equation of GDP is considered as an identity equation. This must hold all the time. Many
economists and policy makes from the government are responsible for setting economic policies
that ensures the efficient allocation of resou rces. Hence, letting Y be GDP,
Y= C + I + G + NX
as discussed above, consumption consists of goods availed in different forms. We now tackle the
second component of the GDP equation, Investment. Mankiw (2010) explained that is it the
availing of goods to be used in the future. Like consumption, it is also categorized into three
subcategories which known as business fixed investment, residential fixed investment, and
inventory investment. Bought property, plant and equipment by firms are considered business fixed
investment. When households purchase housing, this type of investment is called residential
investment. The increase in the good inventories of firms is called inventory investment. Capital
can also be created through investment. Moreover, another c omponent of the GDP equation is
called Government Expenditure. These refers to the purchase of goods and services by the
government, may it be federal, state, or local governments. Government spends for the military,
public infrastructures like government buildings, public schools, public hospitals, highways,
bridges, drainage canals, and services provided by the government. Lastly, Net Exports is the
measure of value of goods and services that other foreign countries buy from a local country; also
known as exports, minus the value of goods a local country buys from foreign countries, also
known as imports. Net exports are positive when the value of exports is greater that imports. They
are negative when imports exceed exports. Trade balance is present when exports and imports are
equal.

Sava Maxim Group: 1737 A.S.E. 2017
4
Econometrics
3. METHODOLOGY
Presentation of Data
This empirical study made use of data containing values for final consumption expenditure, GDP
per capita, UE countries (28). The data was obtained fr om the online database of the EuroStat.
Table 2. UE 28 countries with relative GDP expressed in millions of euros and Final
consumption expenditure.
Country GDP milli on
euros Final
consumption
expenditure
Belgium 382.692,0 299.464,0
Bulgaria 39.940,3 31.951,5
Czech Republic 149.491,1 106.677,2
Denmark 248.974,8 191.741,1
Germany (until 1990 former territory of the
FRG) 2.737.600,0 2.104.460,0
Estonia 18.613,4 13.390,9
Ireland 164.049,8 133244,4
Greece 182.054,2 163.173,6
Spain 1.022.988,0 811.611,0
France 2.059.852,0 1.697.194,0
Croatia 43.127,9 34.869,8
Italy 1.560.023,8 1.252.663,3
Cyprus 16.503,7 14.526,3
Latvia 23.372,1 18.466,1
Lithuania 34.631,2 27.900,7
Luxembourg 45.478,2 22.189,3
Hungary 97.948,0 72.877,9
Malta 7.262,6 5.708,1
Netherlands 602.658,0 443.784,0
Austria 313.066,9 232.343,4
Poland 389.695,1 307.238,6
Portugal 165.690,0 138.406,4
Romania 142.245,1 110.441,3
Slovenia 35.274,9 26.880,6
Slovakia 72.134,1 54.393,1
Finland 193.443,0 159.158,0
Sweden 420.849,1 319.441,9
United Kingdom 1.899.098,0 1.665.154,4
Source: http://ec.europa.eu/eurostat/cache/metadata/en/hbs_esms.htm

Sava Maxim Group: 1737 A.S.E. 2017
5
Econometrics
Source: author’s calculations
National accounts are a coherent and consistent set of macroeconomic indicators, which provide
an overall picture of the economic situation and are widely used for economic analysis and
forecasting, policy design and policy making. Eurostat publishes annual and quarterly national
accounts, annual and quarterly sector accounts as well as supply, use and input -output tables,
which are each prese nted with associated metadata.
Annual national accounts are compiled in accordance with the European System of Accounts –
ESA 1995 (Council Regulation 2223/96). Annex B of the Regulation consists of a comprehensive
list of the variables to be transmitted f or Community purposes within specified time limits. This
transmission programme has been updated by Regulation (EC) N° 1392/2007 of the European
Parliament and of the Council.
Meanwhile, the ESA95 has been reviewed to bring national accounts in the Europea n Union, in
line with new economic environment, advances in methodological research and needs of users
and the updated national accounts framework at the international level, the SNA 2008.
The revisions are reflected in an updated Regulation of the Europea n Parliament and of the
Council on the European system of national and regional accounts in the European Union of 2010
(ESA 2010). The associated transmission programme is also updated and data transmissions in
accordance with ESA 2010 are compulso ry from September 2014 onwards.
Please note, nama will contain the final ESA 95 data transmission from countries, up to 2014 Q2
and will be received until mid September 2014. After this date, ESA 95 data will remain on
Eurobase for analytical purposes. ESA 2010 da ta will be published in a new dedicated database
from September 2014 onwards, called nama10. As countries transmit their data throughout
September 2014, nama10 will run parallel to the existing dataset published in ESA 95, called
nama.
Further information on the transition from ESA 95 to ESA 2010 is presented on the Eurostat
website. There is also a document with information on the transition to ESA 2010 and related data
particularities in the dedicated section under "latest news" that will be regularly upd ated.
The domain consists of the following collections: y = 1,2341x + 5732,7
R² = 0,9963
0,0500.000,01.000.000,01.500.000,02.000.000,02.500.000,03.000.000,0
0,0 500.000,0 1.000.000,0 1.500.000,0 2.000.000,0 2.500.000,0GDP
FCEA Regression Analysis on the Relationship of Final Consumption
Expenditure with GDP (million euros)

Sava Maxim Group: 1737 A.S.E. 2017
6
Econometrics GDP and main aggregates . The data are recorded at current and constant prices and include the
corresponding implicit price indices.
Final consumption aggregates , including the split into household and government consumption.
The data are recorded at current and constant prices and include the corresponding implicit price
indices.
Income, saving and net lending / net borrowing at current prices . Disposable income is also
shown in real terms.
Exports and imports by Member States of the EU/third countries . The data are recorded at current
and constant prices and include the corresponding implicit price indices.
Breakdowns of gross value added, compensation of employees, wages and salaries, operating
surplus, employment (domestic scope), gross fixed capital formation (GFCF) and fixed assets and
other main aggregates by industry; investment by products and household final consumption
expenditure by consumption purposes (COICOP). The data are recorded at cu rrent and constant
prices and include the corresponding implicit price indices.
Auxiliary indicators : Population and employment national data, purchasing power parities,
contributions to GDP growth, labour productivity, unit labour cost and GDP per capita.
Geographical entities covered are the European Union, the euro area, EU Member States,
Candidate Countries, EFTA countries, US, Japan and possibly other countries on an ad -hoc basis.
The data are published:
– in ECU/euro, in national currencies (including euro converted from former national currencies
using the irrevocably fixed rate for all years) and in Purchasing Power Standards (PPS);
– at current prices and in volume terms;
– Population and employment are measured in persons. Employment is also measur ed in total
hours worked.
Data sources: National Statistical Institutes
3.1 LINEAR REGRESION MODEL
To build a linear regression model, we have defined the final consumption as independent variable,
while the value of GDP was considered a dependant (resulta nt) variable. Thus, the regression
model can be written under the following mathematical expression: GDP = a + b · CF
From the econometric point of view, the considered model must include the residual component
too, seen as a representation of the differe nces that occur between the theoretically determined
values and those measured in the real economy.
GDP = a + b · CF + u
where:
GDP = Gross Domestic Product → dependant variable;
CF = Final Consumption → independent variable;
a, b → parameters of the r egression model;
u → the residual variable.
Dependent variable: Final Consumption Expenditure, it measures the market value of all goods
and services, it is in EURO€ .
Independent variable:
GDP : the total income of a country expressed in millions of euros.
Since there are three independent variables and one dependent variable, the model is a multiple
regression model. The model will be estimated with the given equation:
y = 1,2341x + 5732,7

Sava Maxim Group: 1737 A.S.E. 2017
7
Econometrics 3.2 Discriptive Statistics
Table 3. Discriptive statistic elements.

Source: author ’s calcualtion s
The obtained results of the dicriptive statistics shows an the European Union average GDP is
around 466741, 33 millions of euro. This data are not caracteristic for the real situation in EU,
where the major part of the countries are a lot under this numbers. This sitution is due to some
countries like Ger many and UK, that have a GDP much bigger than the rest of the EU countries.
This observation is confirmed by the MEDIAN that is at 164869,9 and is more representative for
the analysed data. The maximum of the EU countries for GDP is in Germany and the mini mum is
in Malta. This is because GDP is not expresed per capita whre surely will be a whole different
situation.
3.4 The linear regression model analysis
Table 4. Regression Analysis for coefficients of correlation
Regression Statistics
Mulitple R 0,998165143
R Square 0,996333653
Adjusted R Square 0,996192639
Standard Error 44296,07542
Observations 28
Source: author ’s calcualtion s
R squared measures the goodness a fit of the model. The nearer it is to 1 the better it is. R Square
= 0.9963 this means that these three variables explain 99% of the variation in the dependent
variable that is final consumer expenditure.

Sava Maxim Group: 1737 A.S.E. 2017
8
Econometrics Table 5. Data of Regression Analysis for Residual and Regression
ANOVA
df SS MS F Significance F
Regression 1 1,38636E+13 1,38636E+13 7065,526941 3,35781E -33
Residual 26 51015699744 1962142298
Total 27 1,39146E+13
Source: author ’s calcualtion s
Our degrees of freedom are k=1
P value is = 3,35781E -33
Also, the viability of this regression model is confirmed by the values of the F – statistic test
(7065,526941 – far above the table level considered to be a mark in viability analysis for
econometric models).
Table 6. Data of Regressi on Analysis of coefficient X1

Coefficien
ts Standard
Error t Stat P Value Lower
95% Upper
95% Lower
95,0% Upper
95,0%
Y
Interc
ept 5732,676
436 10007,80
784 0,572820
395 0,571687
356 –
14838,66
719 26304,02
006 –
14838,66
719 26304,02
006
X 1 1,234134
172 0,014682
165 84,05668
885 3,35781E
-33 1,203954
549 1,264313
794 1,203954
549 1,264313
794
Source: author ’s calcualtion s
From the coefficient we can see that there is a positive relationship between GDP and final
consumption expenditure, the dependent variable. For each unit increase in GDP there is an
increase in final consumption expenditure by 5732 units. It has a negative relatio nship between
final consumption expenditure.
In order for us to accept that there is a relationship between the dependent variable and the
independent variable w e need to see the P values for each variable. The P -value has to be lower
than 0.025 for each variable. GDP P value= 0,571687356 that is smaller than 0.025 so the relation
is not significant.
3.5 Test for Auto -correlation
According to Rufino (2013), auto correlation is endemic in a time series data. Gujarati &Porter
(2009), autocorrelation is defined as, “correlation between members of series of observations
ordered in time [as in time series data] or space [as in cross -section data].” The following are th e
reasons why autocorrelation occurs:
a. Inertia is defined as sluggishness. This is a danger for time series data such as GNP, price
indexes, production, employment, and unemployment which undergo business cycle. Recession
and recovery affects the data of time series.
b. Specification Bias: Excluded Variable Case and Incorrect Functional Form that causes an over
or under estimation in the values
c. Cobweb Phenomenon happens when supply reacts to price with a lag of one time period in the
agricultural sector.

Sava Maxim Group: 1737 A.S.E. 2017
9
Econometrics e. Lags – this is present in a time series of consumption expenditure on income. The variable in the
current period is depended on the past period.
f. Manipulation of data – data is usually manipulated in an empirical analysis.
For example, in a quarterly data, such data is a result of thr ee monthly data which are added and
divided into 3. The procedure of averaging makes the data smooth.
 Durbin – Watson Statistic
The Durbin -Watson d statistic test is considered as the “most celebrated” test in looking for
autocorrelation in a model. It is r eferred to as the ratio of the sum of squared differences in
successive residuals to the Residual Sum of Squares (RSS). D -statistic is based on the residuals
thus it offers a great advantage. This is the reason why Durbin -Watson is included in the table
where 𝑅2, t and F are presented. On the other hand, there are certain assumptions that have to be
noted about d statistic:
a. The intercept term is included in the regression model.
b. The independent variables and the X are non stochastic.
c. The first -order autoregressive scheme generates the disturbances.
d. There is an assumption that the error term 𝑢𝑡 is distributed.
e. Lagged values of the regress and as one of the independent variables are not included in
the regression model.
f. The observations in the data a re complete.
Using the software Microsoft Excel, testing for the Durbin -Watson should result to a d value that is
between 0 and 4. Having a d that is equal to 2 would mean that there is no first order correlation. If
the d value is equal or near 0, there i s positive autocorrelation.
Sum of squared difference of residuals 82041568659
Sum of squared residuals 51015699744

Durbin Watson Statistic 1,608

As seen, the d value of the Durbin -Watson is at 1,608 . According to Gujarati & Porter(2009), since
the number of observation is 28 given k’=1 , the Lower is 1.203 and Upper is 1.264, the model is in
the zone of indecision. The data is in the tails.

 The coefficient of correlation is 1 and R squared value is 0.99, therefore we have an
multicollinear ity.

Sava Maxim Group: 1737 A.S.E. 2017
10
Econometrics Table 7. The independent variable vs. residue diagram

Source: author ’s calcualtion s
The graph leads to the conclusion, by the shape of the cloud of points, that there is no correlation
between independent variables x and residue, that we can say that the model is well chosen.

3.6 The Normality test
Normality Test Score C.V. P-Value Pass? 5,0%
Jarque -Bera 23,01 5,99 0,0% FALSE
Shapiro -Wilk 0,64 #N/A 0,0% FALSE
Doornik Chi –
Square 54,17 5,99 0,0% FALSE

Conclusions
Expenditure on consumption contributes as a portion of the Gross Domestic Product that is also
known as the income of the country. Consumption is established to be very influential in
evaluating the growth of a country; also consumption is a function of national income.
The aim of this paper was to define how the final consumption expenditure affect an important
economic factor as GDP. To test the relationship is needed a statistical analysis to determine how
these variables affect the overall consumption of countries of EU community gathering data from
all 28 countries. Model represent one independent variable: final consumption expenditure and one
dependent variable which is
GDP. Paper shows how certain economic activities have e impact in hoe a country spends for final
consumption. Regression indicates that with the increase of GDP results to increase on final
consumption expenditure. For each unit increase in GDP there is an increase in final consumption
expenditure by 5732 units.

-200000-150000-100000-50000050000100000150000
0 500000 1000000 1500000 2000000 2500000 3000000Residual plot

Sava Maxim Group: 1737 A.S.E. 2017
11
Econometrics References
 CIA World Factbook. (n .d.). Central Intelligence Agency. Retrieved August 21, 2013,
from:https://www.cia.gov/library/publications/the -world -factbook/
 Data | The World Bank. (n.d.). Data | The World Bank. Retrieved August 19, 2013,
from:http://data.worldbank.org/
 Doornick -Hansen Test for Multivariate Normality. (n.d.). Retrieved April 11, 2013
fromhttp://artax.karlin.mff.cuni.cz/r -help/library/asbio/html/DH.test.html.
 Friedman, Milton. (1956). A theory of the consumption function. Princeton: Princeton
University Press/
 Gassoumis , Zachary. (2012) “A negative coefficient for a constant in a linear
regression?”Question answered in Research Gate. Retrived August 25, 2013
fromhttps://www.researchgate.net/post/A_negative_coefficient_for_a_constant_in_a_line
ar _regression
 Reid, Alastair . “Daedalus.” Handout given by Professor Vicente Groyon, De La Salle
University,Malate, Taft. May 28, 2013. pdf.
 Gujarati, D. N., & Porter, D. C. (2009) Basic econometrics (5th ed.). New York: McGraw –
Hill/Irwin.
 http://ec.europa.eu/eurostat/cache/metadata/en/nama_esms.htm#meta_update1428572
830705
 http://ec.europa.eu/eurostat/ca che/metadata/en/hbs_esms.htm
 http://ec.europa.eu/eurostat/data/metadata/metadata -structure